Post Graduate Program in AI and Machine Learning

This pillar offers an in-depth exploration of key machine learning techniques, including clustering, regression, classification, feature engineering, and time series forecasting. With a strong focus on hands-on learning, the module equips learners with the tools and insights needed to design, evaluate, and deploy robust ML solutions across diverse business contexts.

Module 6: Clustering and Principal Component Analysis

Explore unsupervised learning techniques like K-means and hierarchical clustering, while learning how dimensionality reduction through PCA can simplify complex datasets for more effective analysis.

Module 7: Linear and Multiple Regressions

Learn how to model relationships between variables using simple and multiple linear regression, and evaluate model performance with key metrics like R² and RMSE for informed, data-driven predictions.

Module 8: Feature Engineering and Overfitting

Explore methods to refine data, including handling missing values, encoding categorical variables, and scaling features, while ensuring models avoid overfitting or underfitting for optimal predictive performance.

Module 9: Model Selection and Regularization Optimize predictive performance through techniques such as cross-validation and grid search for fine-tuning models, while Lasso and Ridge regression help control complexity and prevent overfitting.

Module 10: Time Series Analysis and Forecasting

Deep dive into techniques to identify patterns in sequential data, capturing trends and seasonality while leveraging models like ARIMA and Exponential Smoothing for accurate future predictions.

Module 11: Practical Application II

Hands-on learning through case studies and mini projects, integrating exploratory data analysis (EDA) with modeling pipelines to apply theoretical concepts in real-world scenarios.

Module 12: Classification and k-Nearest Neighbors

Explore binary classification and the k-NN algorithm, leveraging distance metrics for pattern recognition while emphasizing model evaluation to ensure accurate and reliable predictions.

Module 13: Logistic Regression

Delve into the mathematical foundation of classification, using the sigmoid function for probability estimation while employing evaluation metrics like the confusion matrix and ROC-AUC to assess model performance.

Module 14: Decision Trees

Understand the CART algorithm for structured decision-making, utilizing measures like Gini Index and Entropy to optimize splits while addressing overfitting challenges to enhance model reliability.

Module 15: Gradient Descent and Optimization

Learn to refine model accuracy by minimizing cost functions, exploring batch and stochastic gradient descent techniques to optimize learning efficiency and convergence speed.

Module 16: Classifying Nonlinear Features

Explore advanced techniques like polynomial features and kernel methods to transform data, enabling models to capture complex relationships beyond linear separability for improved classification accuracy.

Module 17: Practical Application III

Prepare for capstone projects through structured assignments, guiding learners in executing comprehensive machine learning projects that consolidate their skills and knowledge.

These advanced modules delves into cutting-edge techniques in deep learning, generative AI, and recommendation systems. Participants will design and deploy robust ML solutions, fine-tune neural networks, and explore LLM-powered applications through hands-on projects. The capstone project ties it all together with real-world problem solving and implementation.

Module 18: Natural Language Processing

Learn the foundational techniques for extracting insights from text, equipping learners with the skills to process, analyze, and structure language data for effective machine learning applications.

Module 19: Recommendation System

Explore intelligent methods for personalized suggestions, leveraging user preferences and data-driven techniques to enhance relevance and accuracy in predictive recommendations.

Module 20: Ensemble Techniques

Get introduced to advanced methods that combine multiple models to enhance predictive accuracy and robustness, leveraging approaches like bagging and boosting to create powerful algorithms for superior decision-making.

Module 21: Deep Neural Networks I

Introduction to the foundational concepts of neural architectures, exploring structured learning through layered networks and efficient signal propagation to enable complex pattern recognition and decision-making.

Module 22: Deep Neural Networks II

Get exposed to the advanced learning techniques, refining model accuracy through backpropagation while exploring the power of convolutional neural networks (CNNs) and real-world applications of deep learning.

Module 23: Introduction to Generative AI

Master the fundamentals of AI-driven content creation, covering generative models, GANs, and the transformative impact of large language models (LLMs) and deep learning architectures.

Module 24: Capstone Project

The Capstone Project guides learners through the end-to-end execution of a real-world data science challenge, from problem scoping and exploratory analysis to model development and deployment, demonstrating applied expertise.

Assignments are graded by industry practitioners who support participants through their learning journey, and/or by the Emeritus grading team.

Late submissions are accepted for up to one week after the program end date. Assignments not submitted by the original due date are marked late and may affect your program completion standing.

The program includes a 6-month IIMJobs Pro membership with access to job insights, recruiter actions, and profile boosts, along with an AI-powered resume builder and career prep modules covering LinkedIn optimization, job navigation, and interview preparation. The program is designed to support positive career outcomes but does not guarantee placement.

Participants receive a verified smart digital certificate from UC Berkeley Executive Education upon successful program completion. The certificate can be shared on LinkedIn, included in a resume or cover letter, and presented to current or prospective employers. This program does not grant academic credit or a degree from UC Berkeley.

You will have access to the online learning platform and all the videos and program materials for 12 months following the program end date. Access to the learning platform is restricted to registered participants per the terms of agreement.

Yes, the qualifying mark is 80%.

Plan for 15–20 hours per week. A typical week includes up to 2 hours of Berkeley faculty recorded lectures, 3 hours of live domain expert sessions, 3–5 hours of self-study and practice, and 5–7 hours of assignments. Weekly hours may vary based on topic complexity and individual proficiency. Note: The above timing aren't fixed and can change basis the topics covered.

Participants get hands-on access to 25+ tools and libraries through sandboxed virtual cloud labs. These include PyTorch, TensorFlow, Keras, Scikit-learn, Pandas, NumPy, Matplotlib, Seaborn, Jupyter, LangChain, Hugging Face, LLaMA (Meta), OpenAI, OpenCV, and Streamlit, among others.

Yes. The program is designed for working professionals and runs fully online. The flexible format pre-recorded Berkeley faculty lectures combined with weekend live sessions allows participants to learn without disrupting their professional commitments.

The program requires a basic knowledge of math and exposure to programming. It is suited for data scientists, software engineers, business analysts, and product managers looking to build or advance practical AI and ML skills. A highly advanced engineering background is not required.